Banks worldwide are investing in and have successfully deployed internal GenAI applications. However, implementing customer-facing applications remains a significant challenge despite their enormous potential to transform banking experiences. McKinsey estimates GenAI could unlock $200-$340 billion annually in banking. This article explores how banks can bridge this gap to transform customer experiences while maintaining security and compliance.

The internal-external GenAI divide

Banks have focused on internal GenAI applications for good reasons: they minimize risk through employee verification, enhance regulatory oversight and ensure secure data management. This mirrors broader industry trends identified in the MIT Technology Review Insights and HCLTech report, where only 23% of business executives felt highly capable of addressing user adoption and trust risks. However, this cautious approach has created a disconnect — while 87% of banking leaders report GenAI investments, few have successfully deployed customer-facing applications.

Key challenges

• Regulatory and compliance risks: Financial regulations like GDPR, SOC 2, PCI DSS and the EU AI Act impose strict requirements on AI-driven customer interactions
• Trust and transparency issues: J.D. Power reports only 27% of consumers trust AI for financial advice, highlighting concerns about accuracy and transparency
• Security vulnerabilities: According to HCLTech's 2024 Cyber Resilience Report, financial services face sophisticated threats including, prompt injection attacks and AI-powered phishing that specifically target GenAI systems
• Integration complexity: Many banks operate on legacy infrastructure not designed for modern GenAI architectures
• Liability concerns: Determining responsibility for AI-generated errors remains unresolved, slowing adoption of fully automated customer-facing solutions

Assessing your GenAI readiness

Before deploying customer-facing GenAI applications, banking leaders should evaluate their readiness across four dimensions: 

Data and AI infrastructure

• Do you have high-quality data to train and refine GenAI models?
• Can your infrastructure support real-time GenAI customer interactions?
• Will GenAI applications integrate seamlessly with your legacy systems?

Security and compliance

• Have you implemented HCLTech's Responsible AI framework to ensure regulatory compliance?
• What measures protect customer data in GenAI interactions?
• Do you have robust audit mechanisms for GenAI decisions?

Customer trust and experience

• How do you ensure GenAI responses are accurate and trustworthy?
• What escalation paths exist from AI to human agents?
• Have you educated customers about GenAI's role in their banking experience?

Operational readiness

• Is there a clear governance framework for GenAI decision-making?
• Do you have the talent to maintain customer-facing GenAI systems?
• How will you manage organizational change during deployment?

Building the bridge: From internal to customer-facing GenAI

Banks should take a phased approach while addressing key challenges to ensure a smooth, secure and effective transition. At HCLTech, we have been building expertise offering a structured approach to transitioning GenAI from internal to customer-facing applications:

1. Start with low-risk internal applications

Begin by implementing GenAI for knowledge management, report generation and risk assessment. These internal use cases develop crucial expertise while minimizing exposure.

2. Implement responsible AI governance

Establishing a governance framework is challenging, but it is crucial for banks to integrate GenAI into customer-facing applications. HCLTech's Responsible AI capability has been helpful to clients in establishing comprehensive governance frameworks addressing:

• Data quality and privacy protection
• Regulatory compliance monitoring
• Bias detection and fairness assurance
• Explainability for transparent decision-making
• Advanced cybersecurity measures aligned with NIST's AI Risk Management Framework

3. Deploy hybrid AI-human models

A hybrid approach ensures AI assists human agents while learning from real customer interactions:

• GenAI augments human agents rather than replacing them
• Routine queries are automated, while complex cases escalate to humans
• Performance metrics track accuracy before scaling automation

4. Run controlled customer pilots

Testing GenAI with a small customer base to refine the model before full-scale deployment. We've observed this in practice with solutions like HCLTech's AI and Cloud Native Labs that provide sandbox environments for:

• Limited rollouts to specific customer segments
• Real-time feedback collection and sentiment analysis
• A/B testing to measure impact against traditional service channels

The future of customer experience in banking

GenAI offers banks significant opportunities to enhance customer experiences; however, successful implementation requires a thoughtful and strategic approach. Banks should transition from internal to customer-facing applications while maintaining compliance, security and trust. Banks can ensure a smooth and effective transition by starting with low-risk internal applications, establishing responsible AI governance, deploying hybrid AI-human models and running controlled customer pilots. Combining the strengths of both human and AI services allows banks to create seamless experiences that meet customer needs and expectations.

The future of banking lies in leveraging GenAI to provide innovative, efficient and secure solutions that transform customer interactions and drive growth.